The world is a complex place — not only because of its intricate networks, but also for the intertwined mechanical and human phenomena that take place within it. Among other tasks, designers work towards recognising and describing interactions between users and systems, aspiring to understand the relationships between rational and non-rational systems. Designing novel products and services through a human-centric approach enables an accurate understanding of users as well as the context and the environment in which they operate. Such an approach, instilled throughout the design process, addresses various human issues so as to facilitate the product and/or service’s adoption and recurring use. As a crucial design issue, user search promotes accurate definitions and representations of users’ environments, situations, contexts, and interactions — a depiction of actual users.
Representing users — environment, context, and situations
There are many available tools to carry out a user search phase. These tools make it possible to define and depict users, to identify their activities and tasks, to understand their interactions with the various elements that form their environment at a given time but also as time passes. Personas, market segmentation, empathy maps, “jobs to be done”,courses, and so on — everyone has their own processes. The common goal of such tools is to make user models so as to design components and therefore build successful experiences.
Questions like this one crop up in many projects: how can we accurately map, in a given context, the behaviours of a large, diverse group of people?
For an interface relating to a job description, for instance, users are markedly targeted and have similar job experience standards. Even if their individual experience standards vary — culturally, cognitively, relating to handicap, and so on — they are still constrained by specific goals pertaining to their work activity.
Let us now imagine an audience as wide and diverse as we have encountered so far — the entire population of France! Developing a project for a French bank, we had to look closely at daily money management in order to dream up new integrated services in the mobile banking space. It is worth noting that almost 99 percent of the French population has a bank account, so every French person is potentially an online banking user. This realisation brought up the issue of how to represent the French population in regards to daily money management.
The granularity of tools such as personas turned out to be too fine to cover a significant portion of the population in a relevant, pinpointed way. Used correctly, personas are perfectly appropriate when designing a product or a service, helping to concentrate on future users’ issues and expectations. But personas did not fit our case — too fine-grained to describe such a large group of people. A persona is made up of motivations, values, frustrations, desires, skills, and so on. It carries its own intrinsic idiosyncrasies and is partly conditioned by its environment — the persona describes a person but not the elements that make up its way of thinking, its behaviour, the way it responds to a situation or a usage mode.
Market segmentation, a rough categorisation of individuals, works in a specific way defined by such marketing attributes as demography, economics, consumption style, behaviour regarding a given product, and so on. This tool and its results closely depend on whatever data was selected and articulated. For very concrete projects related to usage study, we soon realised that usages could fluctuate widely within one segment. Market segmentation is well suited to a vast array of individuals but neither sufficiently accurate nor relevant to define a usage.
We therefore thought up a transitional tool to represent users and depict behaviour and practice patterns that influence a usage.
Usage models: the influence of systems over behaviours.
“Man only ever lives in a permanent relationship with an environment.”(Fraisse, 1968)
Understanding usage models first entails taking notice of systems that might affect human behaviour — whatever is specific to a given person, such as his or her knowledge and mental states — and environment-related elements such as economic contexts and and social groups. These elements, both intrinsic and extrinsic to the person, act on his or her perceptions, decisions, and actions. This is what will trigger behaviours in response to various situations.
By usage we mean a use of something, the action of doing it, or a way of doing it. This could be daily money management, setting up a meeting, getting around while on vacation, and so forth. Here are some examples of major usages triggering various situations that people react to, through practices that are specific to them.
Usage models can be defined as representations of the way people experience situations linked to a usage. We avoid talking about usage linked to a specific object; our focus is on daily usage of which an object might be part of.
Someone who is faced with a situation adopts a certain position and performs practices relative to his or her characteristics, both intrinsic (skills, beliefs, self-confidence, and more) and extrinsic (social, financial, or political capital).
Just like personas, usage models are design tools. Instead of focusing on a person as an isolated individual along with his or her age, skill-set, interests, and more, usage models attempts to understand his or her behaviours and practices when faced with a usage in a given place and time.
This tool is a framework that allows to refocus on personas within the usage we are studying and to understand the causes of their practices when facing this usage. By striving to improve the way we define usages and usage features, our product and service design will gain in accuracy.
Now focusing on one specific usage (e.g. daily personal finance) each person might have one or several usage models, since usage models are neither permanent nor exclusive but flexible. Consciously or not, people might change models over time or at various points of their lives. They can also possess several usage models, such as one in their personal life and another for work, to adjust to switching environments.
Indeed, Serge Proulx, tracing the evolution of usage studies, has emphasised varying logical courses of action, depending on situations or time: “Agents have multiple identities depending on types of situations — spousal, family, work, in public spaces (Lahire, 2001). Agents shift with varying speed between different logical courses of action and various engagement modes following their movements between different types of situations (Thévenot, 2006; Corcuff, 2007).”
Usage Models: Principles and Evolution
Usage models are built upon a usage that implies a large user group. The usage is subsequently dissected, along with its environment, to define behaviour patterns and practices of people who experience this usage.
The first step is to connect with many people around a given subject during interviews, workshops, surveys, and more, attempting to understand their mental frameworks. What does a person see in a usage? How do they experience it? What are their practices? Why do they have such practices?
The goal of the mental framework is to point out common patterns, elements, and characteristics that influence their usage. Empirically, the idea is to let common characteristics emerge from different behaviours and practices linked to one usage and to point out each of these characteristics.
Our process involved writing down all the keywords that cropped up during the user research phase, grouping similar notions. Then we selected characteristics that seemed the most relevant and ascribed them criteria.
This process could be represented by a usage model card containing a summary of the model — a concise explanation of people’s operating mode for this particular usage. We also associated personas to point out that two different personas can share a usage model. These personas were devised during meetings with clients. Next came characteristics directly related to usages — in our example, financial base, account diversity, and so on — and other more general features that are nonetheless influential on practices within this usage, such as people’s vision of the future and self-confidence. Finally, a list of specific needs that people with this type of usage model might encounter.
Long before starting to design concrete elements, working on these usage models allows us to realise the diversity and extent of practices and behaviours related to a usage. Some usages imply many practices while others are much more limited — this gives us a notion of the usage perimeter. Up to us to decide to remain within this perimeter or not, to challenge new practices, or to strengthen others. Similarly to personas, but on a different level, these usage models allow us also to address what has been designed or what will be designed in the future.
An interesting aspect of usage models is their use in scenario design. A persona following a given usage model provides substance to further the relevance of our usage scenarios. A finer understanding of the usage’s context and related users’ reactions can act as a starting point.
We build our usage models empirically, not through existing statistics. Our goal is to confront them to the reality of the terrain so we can either validate, invalidate or reinforce them. Checking and completing them with the client company’s data would also be in order. To what extent can we refine these existing usage models? Did we miss any hidden usage models? Defining and sorting usages should be fed by the client companies’ data whilst respecting data protection laws. Design, data and ethics go hand in hand.
But as Serge Proulx pointed out in an article recapping the evolution of usage studies, it would be unfortunate to fall into a “movement of extreme quantification, possibly eroding the importance of a theory at the cost of a new ideological belief by which data exists in a vacuum, independently from theoretical and methodological models that brought it about in the first place”. (Burrows and Savage, 2014.) It would be unfair to reduce users to bunches of numbers stripped of meaning. What we are trying to achieve is more like a mutual enrichment.
To Conclude
As designers, we must question ourselves about the tools and methods that we use in relation to situations. Our proposal is far from being revolutionary — other disciplines, such as psychology, sociology, information technology, and others rely on similar thought patterns. Our goal is simply to materialise our observations to improve our conceptual design work. This article aims to share our notion of usage models to make progress and find out what others might have achieved in this field, improving the tool as it currently exists as well as its construction.
Initially, usage models were valuable to us as a representation of a vast panel of users. The tool we were lacking had to straddle a wider range than persona models while offering more accuracy than market segmentation, to realistically represent practices of 67 million people.
Our thoughts on granularity of representations are essential. Our tool has allowed us to overcome notions of user tasks and activities to focus on behaviours and practices within a constantly evolving environment. The role and operating mode of usage models when building usage scenarios still need to be developed.
Of course, this tool is not perfect. The material we are working on is not mechanical, and we are not describing the full complexity of reality. We are mainly trying to define a perimeter to guide our design activities. A great deal of intuition occurred while building this tool. Integrating companies’ data to confirm or infirm our hypotheses would be a judicious next step to help us shape these usage models.
We will continue to test and improve this tool, especially as we map new usage scenarios and measure the impact of data in usage models — themes to appear in future articles.
Solène BELLEGO : Lead Service Designer
Thomas NICOT : Senior Service & Strategist Designer
Mikkel KORTSEN : Visual Communication Designer
References
Breviglieri M. (2007), “L’usage, le design et l’architecture. L’éthique professionnelle dans la conception d’un monde habitable” in Les ateliers de la recherche en design (in French), issue 1, p. 53–59 http://gspm.ehess.fr/docannexe.php?id=1495.
Burrows R., Savage M. (2014), “After the crisis? Big Data and the methodological challenges of empirical sociology” in Big Data & Society, 1 April https://doi.org/10.1177/2053951714540280.
Paquienséguy F. (2018), “L’usage, intégré par le design ?” opening speech at an international design research symposium organised by Projekt (EA 7447), “La place de l’usager en Design” (in French), Université de Nîmes, 28–30 March, https://hal.archives-ouvertes.fr/hal-01756008/document.
Proulx S. (2015), “La sociologie des usages, et après ?” in Revue française des sciences de l’information et de la communication, (in French, online), issue 6 https://doi.org/10.4000/rfsic.1230.
Toniolo A. (2009), “Le comportement : entre perception et action, un concept à réhabiliter,” in L’Année psychologique, (in French) issue 109, p. 155–193 https://doi.org/10.4074/S0003503309001067.
